Our ability to navigate, to locate, identify and grasp objects, to judge distances, and to drive vehicles owes much to our having two forward-pointing eyes. Stereoscopic depth perception depends on the slight disparities (position shifts) between the retinal images of the left and right eyes. Knowing how the visual system detects these disparities and uses them to calculate depth is essential for understanding how humans recognize objects and localize them in three-dimensional visual space. It is also essential for finding effective treatment for clinical cases of retinal-correspondence deficits and for improving machine vision algorithms for object recognition, robotics, and visual prosthetic devices. Much has been discovered about the basis for human stereoscopic depth perception and the use of depth cues to infer the shape of objects. The object of this proposal is to update our understanding of how stereo information is computed in light of new evidence about the mechanism responsible. This evidence shows that the way disparities are computed is much more dependent on the detailed spatial properties of the stimuli than previously thought;a visual display may therefore contain only sub-optimal reference stimuli with respect to which the disparity of the target stimulus can be computed. Optimizing the disparity computation therefore requires sophisticated selection of the best reference stimulus consistent with the properties and capacities of the mechanism that carries out the comparison between reference and target stimuli. This project attempts to determine the properties of the mechanism that computes relative disparity and to characterize the reference-stimulus selection process. Psychophysical measures will be used on stimuli that are designed to reveal the basics of stereo computations and also those that will allow a generalization to natural-image objects and real-world tasks.